1. Field of the Invention
This invention relates to a range finding device, and more particularly, it is concerned with the range finding device of a type such that a distance to an object is measured by electrically detecting a relative positional difference between two images in utilization of the distance measurement principle in a base line distance measurement meter. More specifically, the invention relates to a range finding device of a type, in which image scanning signals concerning the two images are obtained by scanning the above-mentioned two images, based on which signals a relative positional difference between the two images may be detected.
2. Description of the Prior Art
There have so far been proposed various types of electric or electronic range finding devices as outlined in the preceding, or automatic focus detecting or adjusting devices for photographic cameras, etc. using such range finding devices. Described more specifically, this range finding device is of such type that two images of an object to be formed by a range finding optical system with a relative positional difference corresponding to the object distance are received by photo-electric light receiving means, and quantities of relative positional difference in these two images are found from outputs of the light receiving means obtained at this time, thereby calculating a distance to the object for the range finding.
For example, according to Japanese Patent Publication No. 48-5733 (Published Feb. 20, 1973, for the invention of "An Automatic Focus Adjusting Device in a Camera" in the name of Canon Kabushiki Kaisha), there is proposed a range finding device in the form of an automatic focus adjusting device, wherein a pair of photo-conductive elements which are so constructed that their resistance values may vary in accordance with positional changes in an image on the light receiving surfaces thereof are juxtaposed, and then images of the one and the same object are formed on these elements by means of a range finding optical system comprising a pair of focusing lenses fixedly disposed at a certain length on the base line so that a distance to the object may be detected by finding a difference in the resistance values between these two elements by utilizing the principle that the quantity of the relative difference in the image forming position on each element corresponding to a distance to the object.
In this disclosed device, however, the photo-conductive element per se is of a very peculiar structure, on account of which there inevitably takes place various inconveniences in using a pair of these elements such that coincidence of the response characteristics in both elements in an ideal state becomes highly difficult, as the result of which precision in the detection becomes deteriorated due to increase in error signal, and false determination in distance will arise in the range detection due to coincidence of the resistance values in both elements in spite of the relative positions of the images on both elements being actually discrepant.
With a view to solving such problem, there has been proposed a range detecting device, or an automatic focus detecting or adjusting device, in which image scanning signals on the two images are obtained by scanning them in utilization of a photo-electric light receiving means, and then by finding a quantity of relative positional difference in these two images from the two image scanning signals, the object distance is calculated, or the focus detection or adjustment of the objective lens system in a camera with respect to the object is automatically attained by utilizing informations concerning a relative positional difference in these two images.
For example, U.S. Pat. No. 3,898,676, (filed Dec. 20, 1973 and issued Aug. 5, 1975, granted to Hosoe et al. for an invention entitled "Distance Detecting Device" assigned to Canon Kabushiki Kaisha) teaches an automatic focus adjusting device of a construction, wherein arrays of photo-sensors are used as the photo-electric light receiving means for receiving the two images, and, by driving these photo-sensor arrays simultaneously, a photo-electric output of each photo-sensor in the arrays is obtained in a timed sequence to thereby scan the two images simultaneously in a purely electrical manner, and image scanning signals to be obtained on these two images at this time are converted to wave form signals through low pass filters, after which these wave form signals are introduced into a phase discriminator to detect a phase difference between these image scanning signals, and a servo-motor is actuated by an output from the phase discriminator to cause an objective lens system to shift along its optical axis, in association with which one of the two images is shifted with respect to the other, whereby "in-focus position" of the objective lens system to the object is determined with a point where the phase difference between the image scanning signals for the two images becomes zero, in other words, a point where the relative positional difference of the two images becomes zero.
Also, according to Laid-Open Japenese Patent Application No. 51-45556 (Laid Open on Apr. 19, 1976 for the invention of "Method and Apparatus for Distance Detection"), there is proposed a method and an apparatus for detecting the distance of an object, which is constructed in such a manner that self-scanning image sensors (a kind of photo-sensor array) are utilized as the photo-electric light receiving means for receiving two images, that the two images are repeatedly scanned by these image sensors, at which time coincidence and non-coincidence of the image scanning signals on the two images to be obtained from the image sensors are detected by means of a coincidence detection circuit, while a timing for commencing scanning of one of the image sensors is varied by a variable delay circuit against a timing for commencing the scanning of the other image sensor, and that the quantity of the relative positional difference of the two images, i.e., the object distance is made known directly from a lagged quantity between the timings for starting the scanning operations by the two image sensors which has been found upon detection of the coincidence of the abovementioned image scanning signals by the abovementioned coincidence detection circuit.
In the methods and devices as disclosed in these U.S. Pat. No. 3,898,676 and Laid-Open Japanese Patent Application No. 51-45556, the two images of an object to be formed by the range finding optical system are scanned purely electrically in utilizaton of photo-sensor arrays or image sensors known as, in particular, Photo-Diode Array, CCD (Charge Coupled Device), or BBD (Bucket Brigade Device), and so on, and the image scanning signals concerning the two images to be obtained at this time are used for the distance detection or focus detection. In particular, since the image is scanned in a purely electrical manner utilizing the photo-sensor arrays or image sensors, accurate signal corresponding exactly to the image pattern can be used, on account of which further improvement in precision of the distance detection or the focus detection can be expected.
However, these methods and apparatuses as have heretofore been proposed contain therein many problems still to be solved such as, for example, concrete method for processing of the abovementioned image scanning signals as one aspect, hence their reduction in practice is far-reaching.
For example, in the device proposed in the abovedescribed U.S. Pat. No. 3,898,676, as already mentioned above, the image scanning signals are converted to the waveform signals by causing them to pass through low pass filters, after which the waveform signals are introduced into the phase discriminator, where detection is conducted to find whether any phase difference has occurred between the two signals. In this case, since the phase discriminator deals with the waveform signals, in particular, the overall construction of the device becomes complicated, and, since its reliability is lacking, it is not possible to accurately detect the phase difference when it is extremely small, on account of which the operation of the phase discriminator as the automatic focussing device is unavoidably inaccurate.
Also, in the device as proposed in the above-discussed Laid-Open Japanese Patent Application No. 51-45556, there is merely adopted a differential amplifier or a combination of the differential amplifier and a comparator as the circuit for detecting coincidence and non-coincidence of the two image scanning signals. Considering, however, that the signals which the circuit deals with are time-sequential signals to be produced, as an output, from the image sensors, it is almost impossible to carry out detection of the coincidence and non-coincidence of the image scanning signals with such simple circuit construction.
Furthermore, in the method and apparatus as proposed here, the scanning start timing of one of the image sensors is caused to vary with respect to the scanning start timing of the other image sensor by the use of a variable delay circuit, and a discrepant quantity between the scanning start timings for these two image sensors is taken as the object distance. However, as has so far been well recognized, in utilizing the self-scanning type image sensors such as the photo-diode array of the charge accumulation type, CCD, or BBD, etc., if the scanning start timing, i.e., the timing for imparting the start pulse, is varied, the integration time, i.e., the effective light receiving time, also varies with the result that the level of the output signal varies. Accordingly, in the proposed method and device, since the scanning start timing of one of the image sensors is caused to vary by the variable delay circuit with respect to the scanning start timing of the other image sensor, the two image scanning signals to be obtained there are resulted from extreme level variations, so that, even if it is attempted to compare these two image scanning signals for detection of coincidence, such is apparently impossible. After all, it is not at all certain to attain accurate distance detection.
On the other hand, quite recently, there has been a proposal of range finding device on the basis of a quite new idea, one of such device is shown, for example, in U.S. Pat. No. 4,004,852 issued Jan. 25, 1977 to Eugene E. Pentecost for an invenion entitled "Integrated Automatic Ranging Device for Optical Instrument". According to this patent, 1st and 2nd images of an object formed by a range finding optical system are scanned for image sensor to obtain image signals on said two images, and after these image signals have been quantized by quantization means, quantized data on M successive elements of the 1st image and quantized data on N(N&gt;M) successive elements of the second image are stored in data storing means, and by the successive comparison of the thus stored M and N quantized data, the location within the 2nd image of one set of M successive elements of the 2nd image which is most similar to the M successive elements of the first image is detected. And from the data of this location, distance between the device and object is determined. According to this proposed range finding device, the above mentioned defects in the conventional range finding devices has been eliminated, the accuracy of range finding is assured so that correct range finding information is obtained. Especially, since analog signals of two images are quantized, data processing of two images becomes easy and accuracy of the detection is improved. Further, digital processing of image data enables to make major parts of the circuit system of the device digitalized, and with the aid of the present technique in the field of semiconductors, the device can be minimized by using integrated circuits which can be easily adapted as an automatic focus detection or adjustment device for a small size optical device such as cameras. By using this device in cameras, high precision focus adjustment can be attained.
However, for realizing such type of device as disclosed in said U.S. patent, various kinds of various improvements are required. For example, in such device, a signal storing type image sensor of photo-diode array, CCD or BBD is used, and as is well-known in this field, this kind of image sensor has a predetermined amount of signals to be stored, so that if an integrated value of the signals represented by "intensity of signal light.times.light receiving time (=integration time)" exceeds the level of the amount of storable signals, there occurs saturation phenomenon, and scan signals properly corresponding to images can no longer be obtained. This necessitates to control integration time of the image sensor. This is generally achieved by, for example, utilizing the time until the data regarding the distance between the device and the object is finally obtained after image scan signals from the image sensor have been processed. However, this time is relatively a long time, and when this processing time is utilized as a part of the integration time of the image sensor to control the integration time of the image sensor, in case of relatively low object brightness, in which the required integration time becomes longer than said processing time, it is possible to attain the satisfactory control of the integration, however, in case of relatively high object brightness, in which the required integration time becomes shorter than said processing time, there happens said saturation phenomenon so that scan signals properly corresponding to the images can be obtained. Especially, when the device is firstly operated or operated only one time, there is no assurance that the signals previously stored in the image sensor would properly correspond to the images, which frequently results in erroneous detection. Consequently, for solving this problem, firstly, it is necessary to provide an improved method.
Further, there is an additional problem to note. In this kind of device, since range determination is made on the basis of the quantized data for the 1st and 2nd images of the object, the accuracy of the range determination depends on output data from quantization means which quantize image signals from the image sensor. If, for example, by any reason, all of the quantized data for the M successive elements of the 1st image become one and the same value, this may be well happened, there would be the case where, in the quantized data for the N successive elements of the 2nd image, the quantized data for the successive elements more than M number become the same as said quantized data for the M successive elements of the 1st image, and in this case, there exist in the 2nd image a plurality of sets of M successive elements identical with the M successive elements of the 1st image so that it is not possible to identify a specified one set. As the result, range determination becomes impossible, so that even if range data is obtained, this is erroneous data on the basis of the erroneous detection. Consequently, it becomes sometimes necessary to provide means to solve this problem.
In this kind of device, on the basis of quantized image data, one set of M successive elements of the 2nd image which is most similar to the M successive elements of the 1st image is detected, and from the location of this one set of M successive elements of the 2nd image, the range is determined. If, during range determination processing, a noise is mixed in the data, it becomes difficult to detect one set of M successive elements of the 2nd image which truly correspond to the M successive elements of the 1st image. In spite of this difficulty, there happens to detect one set of M successive elements of the 2nd image which may not truly correspond to the M successive elements of the 1st image. This is caused by adapting the means to detect one set of M successive elements of the 2nd image which is "most similar" to the M successive elements of the 1st image. For solving this problem, for example, with respect to the one set of M successive elements of the 2nd image which has been deemed to be the most similar one to the M successive elements of the 1st image, it is preferable to evaluate the "degree of similarity", and on the basis of this evaluation, to judge whether said one set would be deemed to be the most similar one set.
Furthermore, when said device is adapted for servo focus control in motion picture cameras, and during the repetitive operations of range finding, if the range data obtained at a certain time of operations is apparently different from the range data obtained at the next operation, it is quite reasonable to make the focus control of the camera on the basis of said range data. On the other hand, when the range data obtained at the next operation is very slightly different from the range data obtained at the previous operation, and focus control is made on the basis of this range data, there happens so-called hunting phenomenon of servo system, which causes trouble in the actual photography. For dealing with these problems, it becomes necessary to control, for example, the output of the device so as to make the focus control system not to respond to the range data.
As mentioned above, in taking various situations into consideration, it is necessary to make many kinds of improvement for realizing said device.